INDEX

Explanations

references to individuals or entities being suspended from their roles or positions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "suspended."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

rics

-0.79

eds

-0.73

ortment

-0.72

 originals

-0.70

swick

-0.70

opher

-0.69

PT

-0.69

 Generations

-0.68

icians

-0.68

atered

-0.67

POSITIVE LOGITS

 suspending

1.14

 suspended

1.13

 suspend

1.04

 indefinitely

1.03

 disbelief

0.98

 suspension

0.98

 suspensions

0.96

 susp

0.92

 lockdown

0.84

©¶æ¥µ

0.84

Activations Density 0.019%

references to individuals or entities being suspended from their roles or positions

instances of the word "suspended."

No Comments

No Known Activations