INDEX

Explanations

terms related to targeting or being targeted

oai_token-act-pair · gpt-3.5-turbo

instances of the word "targeted."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

aos

-0.79

wcsstore

-0.76

itus

-0.73

gall

-0.73

import

-0.71

Reviewer

-0.69

cup

-0.68

rose

-0.66

âĶĢâĶĢâĶĢâĶĢ

-0.66

orian

-0.65

POSITIVE LOGITS

 targeted

3.66

 targeting

2.28

target

2.12

 target

2.01

 targets

2.01

Target

1.45

 Target

1.45

 aimed

1.42

 tailored

1.40

 singled

1.37

Activations Density 0.014%

terms related to targeting or being targeted

instances of the word "targeted."

No Comments

No Known Activations