INDEX

Explanations

phrases related to distance or extent

oai_token-act-pair · gpt-3.5-turbo

the repeated phrase "how far" in various contexts

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

uala

-0.74

emis

-0.70

ritten

-0.69

hler

-0.67

chio

-0.65

onut

-0.64

urrent

-0.64

ortium

-0.63

yre

-0.63

 GOODMAN

-0.62

POSITIVE LOGITS

fetched

1.15

away

0.92

 away

0.86

 flung

0.82

 Ahead

0.77

 distances

0.76

sighted

0.75

thing

0.74

 ahead

0.74

 apart

0.73

Activations Density 0.025%

phrases related to distance or extent

the repeated phrase "how far" in various contexts

No Comments

No Known Activations