INDEX

Explanations

locations or positions described in relation to other objects or landmarks

oai_token-act-pair · gpt-3.5-turbo

prepositions and spatial relationships in descriptions

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

oglu

-0.75

zai

-0.71

odor

-0.71

ober

-0.70

DN

-0.70

Cly

-0.69

atre

-0.68

istani

-0.66

emonic

-0.66

 Assass

-0.65

POSITIVE LOGITS

 foreground

0.98

tyard

0.83

 adjacent

0.81

 beneath

0.79

 adjoining

0.78

 inhabited

0.76

hung

0.73

 beside

0.72

rium

0.72

 corners

0.71

Activations Density 0.123%

locations or positions described in relation to other objects or landmarks

prepositions and spatial relationships in descriptions

No Comments

No Known Activations