INDEX

Explanations

phrases indicating knowledge or lack of knowledge in a subject

oai_token-act-pair · gpt-3.5-turbo

phrases indicating a lack of knowledge or understanding

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

ramid

-0.80

hement

-0.75

odder

-0.74

uably

-0.74

 sidx

-0.73

erate

-0.69

raught

-0.69

nir

-0.69

 Featured

-0.69

 rall

-0.68

POSITIVE LOGITS

 firsthand

0.78

 whereabouts

0.77

 beforehand

0.71

 intimately

0.69

æĿ

0.66

 secret

0.65

ä½

0.65

Orig

0.63

 basics

0.63

LAB

0.62

Activations Density 0.229%

phrases indicating knowledge or lack of knowledge in a subject

phrases indicating a lack of knowledge or understanding

No Comments

No Known Activations