INDEX

Explanations

phrases indicating the speaker's opinion or preference

oai_token-act-pair · gpt-3.5-turbo

phrases that describe distinct sections or elements of experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ternity

-0.82

ilings

-0.67

İĭ

-0.61

ADRA

-0.61

iverpool

-0.61

 kindred

-0.61

artney

-0.60

Topics

-0.60

 actionGroup

-0.59

untled

-0.59

POSITIVE LOGITS

 about

1.11

 ABOUT

0.94

of

0.93

 happens

0.81

 happened

0.76

 thereof

0.76

 regarding

0.75

about

0.75

 though

0.74

 About

0.74

Activations Density 0.059%

phrases indicating the speaker's opinion or preference

phrases that describe distinct sections or elements of experiences

No Comments

No Known Activations