INDEX

Explanations

locations described in a structured format with street names and intersections

oai_token-act-pair · gpt-3.5-turbo

locations and street names

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

xon

-0.76

 worms

-0.71

azeera

-0.71

 posit

-0.67

iously

-0.66

 sheep

-0.66

jad

-0.65

utsche

-0.65

 fide

-0.64

istical

-0.64

POSITIVE LOGITS

 Boulevard

0.89

 Downtown

0.86

 Municipal

0.80

lain

0.79

 âĵĺ

0.76

 Blvd

0.76

 Recreation

0.74

 Floor

0.74

 Historic

0.74

Esc

0.73

Activations Density 0.423%

locations described in a structured format with street names and intersections

locations and street names

No Comments

No Known Activations