INDEX

Explanations

phrases related to legal proceedings, particularly mentions of grand juries

oai_token-act-pair · gpt-3.5-turbo

references to grand jury proceedings and decisions

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

WOR

-0.79

cock

-0.73

comes

-0.72

 Mith

-0.71

gro

-0.69

currency

-0.68

byn

-0.68

BIT

-0.67

owicz

-0.66

ibaba

-0.66

POSITIVE LOGITS

 convened

0.99

 subpoen

0.94

 indicted

0.92

 commissioners

0.90

 hearings

0.87

 subpoena

0.87

 indictment

0.87

 jury

0.84

naire

0.82

 indict

0.79

Activations Density 0.018%

phrases related to legal proceedings, particularly mentions of grand juries

references to grand jury proceedings and decisions

No Comments

No Known Activations