INDEX

Explanations

locations or specific sites mentioned in text

oai_token-act-pair · gpt-3.5-turbo

references to a specific location or site

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

ornia

-0.81

atures

-0.76

 unfocusedRange

-0.73

olyn

-0.68

 Leilan

-0.68

 outgoing

-0.67

sen

-0.66

 french

-0.65

ohan

-0.64

icipated

-0.64

POSITIVE LOGITS

 sites

0.80

 site

0.78

 Site

0.76

 redevelopment

0.76

ologies

0.73

builders

0.71

plan

0.71

 locations

0.70

keeper

0.70

 location

0.70

Activations Density 0.026%

locations or specific sites mentioned in text

references to a specific location or site

No Comments

No Known Activations