INDEX

Explanations

phrases indicating consequences or implications

oai_token-act-pair · gpt-3.5-turbo

the word "which" and its various occurrences in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

athi

-0.71

MENTS

-0.70

BLE

-0.69

stad

-0.60

STE

-0.60

ve

-0.59

Rog

-0.58

atin

-0.58

Behind

-0.57

pay

-0.57

POSITIVE LOGITS

 translates

1.11

 means

1.02

 implies

1.01

 explains

1.00

 begs

1.00

 resulted

0.99

 admittedly

0.98

 brings

0.98

 incidentally

0.97

 happens

0.96

Activations Density 0.113%

phrases indicating consequences or implications

the word "which" and its various occurrences in different contexts

No Comments

No Known Activations