INDEX

Explanations

mentions of locations, specifically ones related to 'Lake' and 'City'

oai_token-act-pair · gpt-3.5-turbo

references to "Salt Lake City" and its related locations

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

binary

-0.90

schild

-0.80

 streng

-0.78

nostic

-0.74

gets

-0.73

akuya

-0.70

mble

-0.69

milo

-0.67

 arrang

-0.66

massive

-0.66

POSITIVE LOGITS

 City

0.98

ois

0.96

 Lake

0.87

 Arabia

0.86

 Valley

0.83

 Falls

0.82

ioxide

0.82

 Ranch

0.81

 Tribune

0.81

Lake

0.80

Activations Density 0.016%

mentions of locations, specifically ones related to 'Lake' and 'City'

references to "Salt Lake City" and its related locations

No Comments

No Known Activations