INDEX

Explanations

references to research outcomes or findings

oai_token-act-pair · gpt-3.5-turbo

mentions of "results."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

resent

-0.75

 actionGroup

-0.70

 Passage

-0.68

pton

-0.66

ti

-0.66

nt

-0.64

afort

-0.63

 Roots

-0.62

agine

-0.60

ModLoader

-0.59

POSITIVE LOGITS

iveness

0.83

iments

0.78

ively

0.76

results

0.76

ivity

0.76

oys

0.74

aries

0.71

 results

0.71

aldehyde

0.69

bringer

0.68

Activations Density 0.047%

references to research outcomes or findings

mentions of "results."

No Comments

No Known Activations