INDEX

Explanations

words related to a geographical feature — specifically, the word "bay" at different activation levels

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "bay" 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

unic

-0.90

inyl

-0.77

PM

-0.73

leton

-0.71

UGC

-0.71

Gener

-0.68

QL

-0.68

Construct

-0.68

GS

-0.68

atum

-0.68

POSITIVE LOGITS

onet

0.93

creen

0.85

bay

0.81

 attendants

0.75

holes

0.72

oused

0.68

eus

0.68

 helm

0.68

hole

0.68

jamin

0.68

Activations Density 0.031%

words related to a geographical feature — specifically, the word "bay" at different activation levels

occurrences of the word "bay" in various contexts

No Comments

No Known Activations