INDEX

Explanations

adjectives related to being slim or having a minimalistic approach

oai_token-act-pair · gpt-3.5-turbo

variations of the word "lean" and related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ilage

-1.10

ilet

-0.83

iles

-0.80

uber

-0.79

ãĤ»

-0.76

protected

-0.75

ourses

-0.74

plain

-0.73

cel

-0.73

vous

-0.73

POSITIVE LOGITS

 toward

0.97

 lean

0.90

 towards

0.90

 shoulders

0.87

 leaning

0.85

 horizont

0.84

 heaviest

0.84

utenant

0.82

 forward

0.79

tip

0.75

Activations Density 0.032%

adjectives related to being slim or having a minimalistic approach

variations of the word "lean" and related terms

No Comments

No Known Activations