INDEX

Explanations

activations in text sections that present technical or official information, possibly related to procedures or authorizations

oai_token-act-pair · gpt-3.5-turbo

occurrences of significant numerical values and dates

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 adolesc

-0.81

xual

-0.76

awaru

-0.72

 estranged

-0.69

 Gardens

-0.66

mur

-0.66

 greenhouse

-0.62

Vik

-0.62

 teen

-0.61

 stomp

-0.61

POSITIVE LOGITS

However

1.03

Traditional

1.00

Unlike

0.99

Furthermore

0.97

Fortunately

0.96

Statistics

0.95

Since

0.95

Generally

0.94

During

0.93

Luckily

0.92

Activations Density 0.330%

activations in text sections that present technical or official information, possibly related to procedures or authorizations

occurrences of significant numerical values and dates

No Comments

No Known Activations