INDEX

Explanations

instances of the word "sat"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "sat" in various contexts

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

ALLY

-0.74

ISO

-0.74

credit

-0.70

Adds

-0.66

#$

-0.65

vernment

-0.64

Operation

-0.63

ever

-0.63

DOS

-0.62

é»Ĵ

-0.61

POSITIVE LOGITS

chel

1.09

anic

0.97

urn

0.91

anism

0.88

olith

0.88

ivas

0.86

 seiz

0.86

iated

0.84

roller

0.82

rooms

0.81

Activations Density 0.018%

instances of the word "sat"

instances of the word "sat" in various contexts

No Comments

No Known Activations