INDEX

Explanations

phrases related to using tools or techniques

oai_token-act-pair · gpt-3.5-turbo

instances of the word "using."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ocratic

-0.69

tv

-0.62

uci

-0.61

imm

-0.59

 aven

-0.59

 flame

-0.59

Ö

-0.59

 nuclear

-0.58

alian

-0.58

 dear

-0.58

POSITIVE LOGITS

 Than

0.90

itialized

0.90

 Issue

0.88

 Ahead

0.86

Beat

0.85

Style

0.85

 Together

0.85

 Again

0.83

 Costs

0.81

 Writing

0.78

Activations Density 0.069%

phrases related to using tools or techniques

instances of the word "using."

No Comments

No Known Activations