INDEX

Explanations

terms related to accusations or claims, particularly of a serious nature

oai_token-act-pair · gpt-3.5-turbo

references to accusations or claims made about individuals or organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

skill

-0.71

tz

-0.69

PRES

-0.67

patch

-0.64

arton

-0.63

spl

-0.62

kin

-0.62

birth

-0.62

perm

-0.62

ï¸

-0.62

POSITIVE LOGITS

 allegations

1.17

 accusations

1.04

 allegation

0.96

 accusation

0.93

 accusing

0.88

 alleges

0.86

 alleging

0.86

 accuser

0.84

 accus

0.77

riott

0.76

Activations Density 0.027%

terms related to accusations or claims, particularly of a serious nature

references to accusations or claims made about individuals or organizations

No Comments

No Known Activations