INDEX

Explanations

verbs followed by adverbs or adjectives describing the degree or quantity of something

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate sufficiency or adequacy

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

lance

-0.66

 Shiva

-0.66

ĨĴ

-0.65

listed

-0.65

________________________________________________________________

-0.64

 Ghana

-0.61

 Lanc

-0.61

 Malta

-0.61

 Mehran

-0.61

ahime

-0.61

POSITIVE LOGITS

 satisfy

1.22

 justify

0.99

 counteract

0.96

 qualify

0.95

 prove

0.92

rouse

0.91

 indulge

0.91

 deserve

0.89

 accommodate

0.88

 make

0.87

Activations Density 0.067%

verbs followed by adverbs or adjectives describing the degree or quantity of something

phrases that indicate sufficiency or adequacy

No Comments

No Known Activations