INDEX

Explanations

words related to hostility and conflict, especially in a negative context

oai_token-act-pair · gpt-3.5-turbo

instances of the word "hostile."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

orah

-0.85

ucket

-0.83

regon

-0.80

frey

-0.79

illon

-0.79

otide

-0.78

ingham

-0.77

annis

-0.77

ulet

-0.75

ainers

-0.74

POSITIVE LOGITS

 takeover

1.00

 hostile

0.90

 retaliation

0.87

 retribution

0.76

 hostility

0.76

 toward

0.76

 invasion

0.75

 towards

0.74

 interference

0.74

 repr

0.73

Activations Density 0.024%

words related to hostility and conflict, especially in a negative context

instances of the word "hostile."

No Comments

No Known Activations