INDEX

Explanations

words related to deviation or straying from a path or standard

oai_token-act-pair · gpt-3.5-turbo

terms related to variation and deviation from norms or standards

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

lain

-1.00

Pain

-0.71

riel

-0.70

HCR

-0.69

te

-0.68

amaz

-0.66

frey

-0.65

roller

-0.65

 Fargo

-0.64

kamp

-0.64

POSITIVE LOGITS

adoes

0.92

ittal

0.87

ippi

0.73

¿½

0.73

 away

0.71

untled

0.70

 erratic

0.68

odox

0.68

 paths

0.68

artifacts

0.66

Activations Density 0.037%

words related to deviation or straying from a path or standard

terms related to variation and deviation from norms or standards

No Comments

No Known Activations