INDEX

Explanations

specific street names and locations from a given dataset

oai_token-act-pair · gpt-3.5-turbo

street names and addresses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

icators

-0.67

otrop

-0.67

 adaptations

-0.66

oan

-0.65

laus

-0.63

 breeds

-0.63

 manuals

-0.62

omsky

-0.62

onom

-0.62

rogram

-0.62

POSITIVE LOGITS

Rd

0.94

 Corridor

0.93

NW

0.92

 Blvd

0.90

 Plaza

0.88

 Parkway

0.88

 intersection

0.87

 corridor

0.86

 Bridge

0.85

 Extension

0.85

Activations Density 0.070%

specific street names and locations from a given dataset

street names and addresses

No Comments

No Known Activations