INDEX

Explanations

mentions or references to knives and related actions

oai_token-act-pair · gpt-3.5-turbo

references to knives and their associated actions, particularly in the context of violence or conflict

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

mberg

-0.80

auga

-0.79

phas

-0.78

rians

-0.70

GROUND

-0.69

 Leban

-0.69

rian

-0.68

oS

-0.68

leep

-0.68

tainment

-0.68

POSITIVE LOGITS

 blade

1.18

 scissors

1.09

 blades

1.06

 knife

1.05

 knives

0.97

knife

0.94

 wielding

0.94

 wounds

0.93

 stabbing

0.92

 stabbed

0.92

Activations Density 0.052%

mentions or references to knives and related actions

references to knives and their associated actions, particularly in the context of violence or conflict

No Comments

No Known Activations