INDEX

Explanations

words related to deception and disguise

oai_token-act-pair · gpt-3.5-turbo

instances of the syllable "gu" in varied contexts

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

cling

-0.76

croft

-0.75

 Spectrum

-0.72

rings

-0.69

hower

-0.69

riad

-0.68

cycle

-0.67

HAEL

-0.65

ŃĶ

-0.64

âĶģ

-0.61

POSITIVE LOGITS

pta

1.16

ilty

1.14

idelines

1.13

errilla

1.09

arding

1.09

cci

1.08

vernment

1.05

arant

1.03

inea

1.01

ests

1.01

Activations Density 0.021%

words related to deception and disguise

instances of the syllable "gu" in varied contexts

No Comments

No Known Activations