INDEX

Explanations

words related to a person named "Till"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Till" and phrases related to time measurement

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

TABLE

-0.74

Premium

-0.70

Jane

-0.67

Loading

-0.67

GU

-0.67

GROUP

-0.66

Es

-0.66

BILITIES

-0.65

Cath

-0.65

iliary

-0.65

POSITIVE LOGITS

ting

1.00

ted

0.90

unda

0.85

stad

0.81

 Till

0.76

 Nadu

0.76

stal

0.74

uminati

0.74

awar

0.73

umi

0.73

Activations Density 0.018%

words related to a person named "Till"

occurrences of the word "Till" and phrases related to time measurement

No Comments

No Known Activations