INDEX

Explanations

legal or criminal involvements in a given text

oai_token-act-pair · gpt-3.5-turbo

instances of the word "involving" and related contexts indicating connections to various subjects or events

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

 teasp

-0.78

imet

-0.76

©¶æ¥µ

-0.76

nery

-0.75

asis

-0.75

nis

-0.74

omet

-0.72

athy

-0.71

ochond

-0.71

owicz

-0.70

POSITIVE LOGITS

 underage

0.64

 regards

0.63

 involving

0.63

RTX

0.60

 minors

0.60

 allegations

0.60

 Curious

0.59

between

0.59

 improper

0.58

 both

0.58

Activations Density 0.063%

legal or criminal involvements in a given text

instances of the word "involving" and related contexts indicating connections to various subjects or events

No Comments

No Known Activations