INDEX

Explanations

entities or concepts related to legal proceedings or political events

oai_token-act-pair · gpt-3.5-turbo

instances of the word "ent" in various contexts, possibly related to entrapment, entertainment, or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

BILITIES

-0.76

 Accountability

-0.74

 Responsibility

-0.73

 Spears

-0.73

å§«

-0.71

å£«

-0.69

BILITY

-0.68

 Jenner

-0.66

strap

-0.64

ãģ®éŃĶ

-0.64

POSITIVE LOGITS

renched

1.02

ourage

1.01

rave

0.99

ailed

0.99

inence

0.97

rench

0.95

ent

0.93

uri

0.91

ailing

0.91

uring

0.90

Activations Density 0.004%

entities or concepts related to legal proceedings or political events

instances of the word "ent" in various contexts, possibly related to entrapment, entertainment, or entities

No Comments

No Known Activations