INDEX

Explanations

phrases related to conflicting situations or disagreement

oai_token-act-pair · gpt-3.5-turbo

references to conflicts and their implications

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

dig

-0.77

 Tube

-0.74

ait

-0.68

Station

-0.64

charg

-0.64

Tube

-0.64

 Tail

-0.63

oyer

-0.63

 Believe

-0.62

kt

-0.61

POSITIVE LOGITS

 conflicts

3.92

 conflict

2.72

 Conflict

2.06

 clashes

1.94

 disputes

1.75

 conflicted

1.73

licts

1.73

 conflicting

1.66

 conflic

1.60

 disagreements

1.53

Activations Density 0.019%

phrases related to conflicting situations or disagreement

references to conflicts and their implications

No Comments

No Known Activations