INDEX

Explanations

references to physical rooms or spaces

oai_token-act-pair · gpt-3.5-turbo

mentions of rooms 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

Uz

-0.67

 retribution

-0.65

HER

-0.64

Iss

-0.63

FTA

-0.59

PB

-0.59

FF

-0.58

 Garc

-0.58

FIN

-0.58

 Accessed

-0.58

POSITIVE LOGITS

room

1.17

rooms

0.95

 rooms

0.93

 room

0.90

 Room

0.88

stairs

0.86

itory

0.85

 upstairs

0.84

 doors

0.82

Room

0.81

Activations Density 0.024%

references to physical rooms or spaces

mentions of rooms in various contexts

No Comments

No Known Activations