INDEX

Explanations

words related to information retrieval or analysis

oai_token-act-pair · gpt-3.5-turbo

actions related to evaluation or assessment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ovie

-0.80

jet

-0.72

athi

-0.72

Bio

-0.71

conn

-0.69

rams

-0.68

oil

-0.67

here

-0.65

script

-0.65

bill

-0.64

POSITIVE LOGITS

 whether

0.80

ially

0.75

ively

0.73

lement

0.70

 thresholds

0.69

 determine

0.68

 determines

0.67

how

0.65

 Lauder

0.65

nda

0.65

Activations Density 0.018%

words related to information retrieval or analysis

actions related to evaluation or assessment

No Comments

No Known Activations