INDEX

Explanations

words related to aggression and conflict

oai_token-act-pair · gpt-3.5-turbo

suffixes related to adjectives and nouns, particularly those indicating states, conditions, or processes

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

DragonMagazine

-0.84

Bloom

-0.77

owler

-0.76

ãĥ¯ãĥ³

-0.70

Answer

-0.69

 teasp

-0.69

vertisement

-0.69

ffield

-0.68

 challeng

-0.68

 nodd

-0.67

POSITIVE LOGITS

ously

0.78

 thereof

0.71

ially

0.70

iously

0.70

aires

0.68

ities

0.66

ITIES

0.66

naires

0.65

ionage

0.63

 fractures

0.62

Activations Density 0.252%

words related to aggression and conflict

suffixes related to adjectives and nouns, particularly those indicating states, conditions, or processes

No Comments

No Known Activations